Skip to content

What this is

Your numbers can improve the benchmark without anyone being able to read your business out of it.

Audits can improve the ranges used in future audits when operators choose to contribute. That depends on a pool that cannot be traced back to an individual business. This page explains how participation works and how those rules are enforced.

The calculator is on the home page under "Let's find the gap." This page explains participation and limitations.

How private evidence contributes to a shared reference

Your business stays out of the record.

  • Names
  • Customer records
  • Invoices

Records describe a segment.

  • Trade and revenue band
  • Region and coverage
  • Percentiles, basis and sample size

Conceptual record shapes, not client observations or current benchmark values. Contribution follows the participation terms below.

The calculator currently uses provisional design-time estimates, not observed data. Read the source disclosure below.

What goes in

Group metrics, without company records.

Inside a benchmark record
Segment
Trade, revenue band, region and sometimes crew coverage
Metric
The measure being compared
Basis
Where the figure came from
Sample size
How many businesses are behind it

Three positions in a distribution. Select one to see what it means.

25thLower quartile

About a quarter of observations fall at or below this point. This is a rank position, not a 25% performance target.

50thMedian

The middle of the distribution: half the observations lie at or below it. It describes the group, not a named company.

75thUpper quartile

About three quarters of observations fall at or below this point. Higher is not necessarily better; that depends on the metric.

Anatomy illustration only. No metric values, sample count or observed dataset are shown.

A benchmark row describes a metric for a group of businesses, defined by trade, revenue band, region and sometimes crew coverage. It includes the 25th, 50th and 75th percentiles, the source of the figure and the number of businesses behind it. It contains no name, address, customer, invoice or figure belonging to one operator.

A band that could only have come from one business is not eligible. The sample size recorded with each row lets you check the size of the group.

What it cannot contain

The benchmark table has no identifying fields.

The record boundary
01Not in the pool

No identity fields.

  • Organisation or client column
  • Company, contact or engagement link
  • Name, address, customer or invoice
02In the pool

A group and its evidence.

  • Segment and metric
  • 25th, 50th and 75th percentiles
  • Source basis and sample size

Adding an identifying field would change the schema. The build guard checks for that change.

A conceptual view of the declared benchmark fields and exclusions, not a security certification.

The benchmark table has no organisation or client column and no link to a company, contact or engagement. Adding any of those fields would require a schema change.

A guard runs on every build and fails if it finds a tenant column or identifying foreign key in the benchmark tables. We added it after writing this page exposed a gap: the rule had been documented but was not enforced.

The estimator on the front page

The estimator uses six figures and does not require an account.

A calculation record, explained
01Recorded when the tool runs

The calculation.

  • Selected segment
  • Six input values
  • Ranges returned
  • Time taken
02Absent from that record

No email gate.

  • No email field
  • No contact
  • No company

A browser-generated identifier distinguishes sessions. It is not joined to an identity unless you later choose to identify yourself.

Illustration of the recorded fields described below. This is not a form and does not run the estimator.

The benchmark tool takes six figures and returns ranges. There is no email gate.

We record the segment, the six input values, the ranges returned and how long the calculation took. The record has no email, contact or company field. A first-party identifier generated in your browser distinguishes sessions. It is not an identity and is never joined to one unless you later choose to tell us who you are.

Use the benchmark calculator

What a refusal means

When the pool is too thin, the tool says so instead of guessing.

Three conceptual response paths

Matched

Published rows match the requested segment.

The requested segment

A matching range

Use the matching basis.

This example uses matching published rows and states their basis and business count.

Broader

No published rows match the requested segment.

A broader segment

The broader basis is disclosed

Widen the comparison, and say so.

If an exact segment is unavailable, the tool may use a broader comparison and disclose that basis.

No range

The tool declines to offer a range.

No supported range offered

No numerical estimate

Refuse the range rather than guess.

A refusal leaves the range unstated. These gaps show where the pool needs more depth.

The basis behind a band

Must be stated.

The number of businesses behind it

The sample size is recorded with each band.

Conceptual comparison paths, not live estimator results. Current ranges are provisional design-time estimates.

Each band must state its basis and the number of businesses behind it before it can be stored. If no published rows match your segment, the tool either widens the comparison and tells you, or declines to give a range.

We track how often this happens and which segments are affected. Those gaps show where the pool needs more depth.

Taking part, and leaving

You can choose not to contribute without changing your audit.

Your audit is the same either way.
01Contribution

Your aggregates may contribute.

  • The data-handling terms come in writing first
  • De-identified aggregates may enter after a delivered audit
02No contribution

You can choose not to contribute.

  • Tell us that your figures should stay out
  • It changes nothing about your audit

The benchmark you are measured against does not depend on you being in it.

An explanation of participation, not a consent form. No preference is recorded on this page.

If we deliver your audit, de-identified aggregates from it may enter the pool. You receive our data-handling terms in writing before you share anything.

Tell us if you do not want your figures to contribute, and we will keep them out. Your audit is unchanged: the benchmark used to measure it does not depend on your participation.

Read how we handle your data

Where the published figures come from

Five published industry sources.

The named source index

  1. ACCA
  2. PHCC
  3. ServiceTitan
  4. Housecall Pro
  5. IBISWorld
Published industry sources named in the methodology. Source attribution, not clients or endorsements.

The estimator's ranges draw on published industry sources: ACCA, PHCC, ServiceTitan, Housecall Pro and IBISWorld. Every figure in a delivered audit cites one of them by name.

The current ranges are provisional design-time estimates rather than observed data, and the tool says so on every result. They are replaced with researched figures before launch.

Benchmark participation · Anchorfield Consulting Group